test: move two tests in serve to gpu-integ-tests

[lucasjia-aws]

Issue #, if available:

Description of changes:

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[github-actions]

Claude finished @lucasjia-aws's task in 28s —— View job

---

PR Review

• Read PR diff
• Verify marker is registered and usage is consistent

Summary: This PR adds the @pytest.mark.gpu_intensive marker to two integration tests (test_recommendation_workflow_end_to_end and test_jumpstart_build_deploy_invoke_cleanup) so they run on scheduled CI rather than PR checks.

The change looks clean and correct:

• The gpu_intensive marker is properly registered in sagemaker-serve/tox.ini:66 ("mark a test as GPU resource intensive (runs on scheduled CI, not PR checks)"), so there are no unknown-marker warnings.
• The marker is applied consistently alongside the existing @pytest.mark.slow_test decorator on both tests, matching the intent of moving GPU-heavy tests off the PR path.
• No functional/behavioral code is touched — this is a test-selection metadata change only.

No issues found. 👍